Oral-History:Chalmers Sherwin

From ETHW

About Chalmers W. Sherwin

Sherwin received his BA from Wheaton College in physics and math in 1937 and his PhD from the University of Chicago in 1940. He was recruited into the Rad Lab in 1941 by Rabi, who liked the fact that Sherwin had done some amateur radio work—it demonstrated a practical turn of mind he valued. At the Rad Lab, Sherwin worked under Bob Bacher, Lee Haworth, and Ted Soller, in indicators, eventually in the circuits group. He considered his most important work to be the fully electronic Plan Position Indicator, the display for ground control approach, and the development of an oscillator capacitor device to compensate for nonlinear scan signal or output shaft.

After the war Sherwin worked for some academic institutions, but also at the Control Systems Lab during the Korean War, as a scientific advisor for the government, and at Aerospace Corporation. Considering Rad Lab’s postwar influence, he notes that the display groups went on to do a great deal of research on TV displays; more broadly, he thinks the Rad Lab management model—director-control, not peer review control, goal-oriented, a direct knowledge of military requirements—had, and deserved, great influence, particularly at the Control Systems Lab, the Charles River project (later Lincoln Labs), Project Wolverine, and the Materials Research Laboratories. He considers it a very good model for applied research, made sure when he was in the government to get government research labs on that model as much as possible. He notes Project Hindsight determined this was the most effective way to develop improved military systems.

About the Interview

Chalmers W. Sherwin: An Interview Conducted by John Bryant, IEEE History Center, 12 June 1991

Interview # 082 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Chalmers W. Sherwin, an oral history conducted in 1991 by John Bryant, IEEE History Center, Piscataway, NJ, USA.

Interview

Interview: Chalmers W. Sherwin

Interviewer: John Bryant

Date: 12 June 1991

Location: Boston, Massachusetts


Family Background and Education

Bryant:

This is the 12th of June 1991. I'm John Bryant interviewing Dr. Chalmers Sherwin for the IEEE Center of History and History Committee Project on Oral History for the MIT Radiation Lab. Dr. Sherwin, could we start with some background, perhaps? Your family, your parents, or why you chose to become a physicist, a scientist?

Sherwin:

Well, that's interesting. There were no scientists in my family. My father was a Presbyterian minister. He died in a drowning accident when I was only five years old. So I was raised by my mother and had no contact with any scientific activity. The turning point in my life occurred when I received, I think at about ten or eleven years of age, an electrical set, which was a box about this big. It had a motor in it. It had some coils and a bit of mercury to make a bouncing spring. And it had some electroplating experiments. They had about 50 experiments in this electrical set. I found it absolutely fascinating! That is what turned me into science. First I was going to be an electrician. That's the first I knew about it. Then I got interested in radio, and with an electrical set I got interested in physics and its applications. As a matter of fact, I recently discovered there's a number of beautiful books put out by various publishers on home experiments for children and how to teach them science at home. I've been getting copies of these books and distributing them to my grandchildren to try and get them interested in science by having them actually do experiments. But that's what got me interested. Then I went to high school and got even more interested. I took the physics courses, chemistry courses, but my decision really to go into science occurred just about the time I entered high school. And it's never wavered since. Then I went to Wheaton College, got a four-year degree in physics and mathematics. That wasn't a very strenuous thing. I left Wheaton College in '37 and went to the University of Chicago, where I had been enamored of Arthur Compton. Arthur Compton got a lot of publicity in the early 'thirties. He, of course, had won the Nobel Prize for his Compton effect work at Washington University, and then he was at Chicago. He was working on cosmic rays. He was going to the tops of mountains, and he was going up in airplanes and sending up balloons and all these exciting things. So there was a lot of public information about it. I was very impressed with Compton. So I wanted to go to the University of Chicago.

I started there in the physics department, graduate school, September 1937. I got through in 1940, in June. Got a Ph.D. in experimental physics with Arthur Dempster, who was one of the great mass spectroscopists. I did work in the mass spectroscopist properties of gases and the transfer of charge. Then I was given a research appointment for nine months at the university and started working on efforts to build a high-intensity uranium ion source because that was the time when there was already interest in nuclear weapon development. The idea was to generate uranium ions and separate them with the mass spectrometer, which is in fact what finally happened. They built those things down at Oak Ridge. Dempster had developed a method of sparks, which was through spark metals. This produces ions, and then you can pull the ions out. He thought it might be an efficient way of getting high currents. I spent about two or three months on that and really made very little progress.

Rad Lab

Recruitment

Sherwin:

In March or May of 1941 I.I. Rabi came through scouting for people. I remember talking to Rabi, and he asked me about my experience. And I said that I really hadn't done much electrical or, electronic work, but I'd been an amateur radio operator for a number of years. "Ah!" he said, "That's what we want! That's the kind of people with a practical turn of mind." In those days in my physics work, we used quadrant electrometers to measure low currents. There were no electronic amplifiers or anything to draw on. It was fascinating. I knew how to set up the quadrant electrometer and have beams of light going across the screen and all that sort of thing. What made me acceptable to Rabi was that I had been in radio. I had a radio station for a while. Within a few weeks, I packed up the whole family, two small children in our 1940 Chevy, and we headed to Boston.

Bryant:

This answers my question as when you were first told about the then-secret subject of radar.

Sherwin:

I don't recall when that was. I think he talked about the general activity in general terms that improved microwave sources had been developed, and they were collecting a bunch of people on an emergency basis to try to develop practical systems based on it. So I knew about the general nature of the work before I came.

Bryant:

A lot of academic physicists had already made the decision. Many of them left their institutions very quickly and went to Cambridge. Some left in mid-term.

Sherwin:

Chicago didn't really need me. I was making $1800 a year there, and the opening salary at Radiation Lab was $3600 a year. That looked awfully good, and of course in those days it was. We lived very well on $3600 a year. We rented an apartment in Belmont for $50 a month or something like that.

Bryant:

In Belmont?

Sherwin:

Yes.

Bryant:

Beautiful area.

Sherwin:

Yes. It was a duplex.

Bryant:

Did anyone give you assistance on finding housing when you came?

Sherwin:

I don't remember. I don't think so. We shopped around. We were given some advice on where to look, and we just went to real estate people.

Indicator Group

Bryant:

You apparently went to work at Radiation Lab about May of '41.

Sherwin:

Yes.

Bryant:

You went right into the indicator group there, I guess. What kind of indoctrination did you get? Or was the employment procedure fairly simple?

Sherwin:

Yes. I don't even remember it. Bob Bacher was the head of the indicator group when I arrived.

Bryant:

It was also the receiver group, wasn't it? They were put together?

Sherwin:

They may have been together as the same group at the same time, yes, before they separated the two. Within a few months, Lee Haworth took over the indicator group, and Bacher took on a division responsibility of some sort. The first problem I worked on was developing a B scan with a magnetic deflection tube. The work that had been done up to then had been mainly with electrostatic deflection tubes, and they had poor focusing properties for intensity. Magnetic focus was much sharper, and you could get higher voltages on them. The problem was to develop deflection systems. The yokes and the amplifiers that would drive them that would allow you to get linear sweeps at various speeds which he controlled.

That turned out to be very complicated because I'd never done anything like that. I discovered I didn't know how to calculate the circuits, so I went and got a book on Laplace transforms. I've never seen them generally used. Engineers use them a lot. You use Laplace transforms to solve linear differential equations, which are otherwise very complicated to solve. Every little differential equation is an exercise in imagination. This one you simply put in the symbols for the different forms of derivatives, the Laplace transform, and then by a simple transformation you're able to get the solution to the equation. Somebody told me about it. So I went and bought one, and I studied up on Laplace transforms, and pretty soon I was able to figure out how to get the type of waveform you need on the coil. But we had a terrible time getting enough speed on these sweeps, enough high speed to get in five-mile range. We ended up actually bringing the current up and then cutting it off and allowing it to oscillate and using the straight portion of the oscillation. The problem with that was that you had to know ahead of time where the transmitters going to go out.

Bryant:

Modulators had that feature, too.

Sherwin:

Yes. So we worked at that. To me it was totally new. I'd never done anything like this before, and I'm sure it was very inefficient. But we worked six days a week. Nobody complained, and everyone worked hard. It was a very exciting period. Some people look back at it as one of the most interesting parts of their careers.

Bryant:

I can think so.

Sherwin:

Yes.

Bryant:

The motivation and the concentration on the specific goals was unprecedented, I guess. And also the group participation, the group effort.

Seminars

Sherwin:

Yes. And they had seminars several times a week.

Bryant:

I'd like to know more about. I've heard a lot about the Hansen Lectures and the Monday evening lectures. I'm not sure how they all sort out. There were some lectures held on a regular basis where the committees controlled the subject matter and the speakers, right?

Sherwin:

I don't remember. All I remember is that they had lectures on the overall systems design and principles of the radar system, about the transmission and how it falls off at 1/R and then how you get 1/R^4 when you get a reflection from a distant source. There were also some lectures on some of the principles of waveguides and how they worked, and how the transmit-receive switches worked. In other words, they were describing the overall system concept to all of the engineers and scientists involved. Someone recently made a comment about the excellent systems engineering that Radiation Lab did. They looked at the whole system, right from the beginning. And they trained everybody to understand the overall picture as well as their own part.

Bryant:

Were some of these seminars held on a regular basis?

Sherwin:

I have the recollection that they were. Once or twice a week there was a one- or two-hour seminar that everyone was expected to attend.

Bryant:

So you were expected to attend?

Sherwin:

Yes. Part of the job.

Management and Information Exchange

Bryant:

Well, that's a people-to-people subject, and people do it, so they have to be brought together. How much interaction did you have with people in other divisions? On a need basis? Or was this informal?

Sherwin:

I just don't recall how we interacted with the other divisions. The main interaction was with the systems people that needed the functions of the display. They had to have certain properties, had to cover certain angular rates and have certain range speeds and certain accuracy of timing. They didn't like the pre-triggering, so we had to develop methods of getting very high step-voltages, very rapidly applied, so that you could get a rapid generation of a current waveform. The interaction with Luis Alvarez was later, that was after a couple of years when they were developing the ground-control approach system. They had a problem of the display which would magnify the vertical direction on the display so they could see accurately how the plane was coming in. The plane is being tracked coming in on a beavertail-type scan, and the plane is supposed to come down on a straight line. If you put that cross-section on a standard PPI, it may be out at 5 miles but only 1,000 feet up. So you'd have a line that was only a very slight angle when it was horizontal. So we worked out a method, which magnified a display in the vertical direction by a factor n, that all the range marks appear now as ellipses rather than circles, but straight lines are still straight lines. So we figured out how to do that. That we did in great discussion with the people. Alvarez's antenna (he had a squeezable waveguide antenna) worked such that the distance of the waveguide motion caused the beam to go back and forth, but it was a nonlinear relationship.

Bryant:

He was changing phase velocity in a different way, that's all.

Sherwin:

Yes, right. Consequently the detection, the drive for the sweep that determined angle, had a nonlinear relationship. We actually manufactured capacitors that rotated with this motion. And we shaped the capacitors so that the capacitance varied.

Bryant:

It was an analog solution.

Sherwin:

Right. An analog solution to an analog problem. Yes. Today we do it differently. But I remember we worked out and built those special capacitors. Don't know where he got them built. But they were mounted right out on the antenna because they had to pick up the motion.

Bryant:

The position motion?

Sherwin:

Right. Then we had oscillators of some sort on that. Then the frequency would change, and somehow or other we measured the frequency. I don't know how we did it. Well, we had to have a frequency discriminator. That was an example where the vertical magnification of the display and the nonlinear take-off were all tailored exactly to that particular radar system.

Bryant:

It's quite a system. I wonder if we could get into a discussion of Radiation Laboratory management and social organization. To what extent did the flow of information follow the channels as defined by the formal organization?

Sherwin:

I don't remember any formal information flow at all.

Bryant:

No?

Sherwin:

The only information was this: You get together with Mr. X over in GCA radar system and get all the specs and get your display to meet what they require. In other words, people were expected to have direct communication with the systems people. Nobody wrote specifications that were approved by some higher authority and followed with directions to the component groups. The total directions were: get together and work out a solution and tell us what happened. The people responsible were the systems managers. They were responsible for the systems to work. If some of the component people did not or were not able to do the job, the systems manager had to solve it some other way.

Bryant:

There had to be objectives, but was there any formal scheduling arrangement that you were aware of?

Sherwin:

Nothing except specific schedules for each individual project. For example, we had to get the MTI ready for the SCR-598 of this prototypes available at a certain date, and then we had to do engineering improvements. I was sent down to GE Bridgeport a couple of times while they were manufacturing them to be sure that they were doing it right, they had some problems. I even went down to Fort Monroe, Virginia once on the SCR-598 because they had some display problems on that. They were testing it then.

Bryant:

You had the authority to ascertain that a trip was needed and make it on your own decision?

Sherwin:

No, as I recall it, trips like that were taken at the request of the systems manager. You could ask to go, and I'm sure they would not have argued about it, but you were not expected to independently go out and get involved in some systems problem. You were expected to work with a systems manager. Is that what other people have said? I'm just curious.

Bryant:

That's very much in keeping. I guess I'm interested in how projects got formulated. Do you have any feel for it? How much things were generated internally or were the Armed Forces introducing their own requirements into it?

Sherwin:

I wasn't involved in that so much. In the components, such as a display, we really depended upon the systems people to define their functions, and it was a responsive role. We didn't invent new systems. We had to have a good PPI display for an X-band airborne radar for ground mapping. We never got involved in the planning of the system: what radar would work and what the sensitivities were. We only got involved after the systems people had decided it was possible. Then they would give specs to the components departments. So we had a supporting role.

Inhouse Work and External Resources

Bryant:

What facilities did you have in-house for your own experimentation? Did you have any experimental work in cathode-ray displays, or did you depend on industry for making what you wanted?

Sherwin:

The indicator group had two parts: the circuits part under Ted Soller. After it was later separated, I had the circuits and Soller, Ted Soller, had the cathode-ray tube displays. He did more work on looking for improved phosphors and looking for daylight displays in which you would cause a change in color of a thin white solid-state material. I've forgotten the name of what they were called, but the idea was to be able to have a display that could be used in full daylight. Soller had conversations with a number of people, including, I remember, Fred Seitz and other people who were leaders in the solid-state physics of the period about what kind of phosphors to get. I don't remember whether we formulated any or whether we simply kept getting samples, but there was a considerable research effort in improved phosphors, mostly for daylight. The only thing we ever used was the green display phosphors that had been developed years ago. I don't think they were ever improved.

Bryant:

Did he have his own cathode-ray tube making facilities to make his own experimental displays?

Sherwin:

Yes, they did. They had vacuums and guns. They wouldn't make complete tubes, but they would make up a vacuum situation in which they would have a gun with a deflection system, and they would do tests with phosphors. But in the circuit end, we didn't have any formal R&D programs. There was a general requirement for making what we called range marks, making decision-timing devices which could put rings on cathode-ray tubes and markers on. We developed that ourselves, but it was very simple, just dividing various types of oscillators and sharpening rectifiers and sharpening devices to produce limited pulses.

Bryant:

How much did you find yourself depending on outside sources for techniques and technology?

Sherwin:

We had very little interaction outside.

Bryant:

Very little need for it?

Sherwin:

Yes.

Bryant:

Commercially available things could be adapted?

Sherwin:

Yes.

Bryant:

I take it the storeroom was pretty well stocked.

Sherwin:

Yes, they had all the standard tubes and capacitors and parts and switches. We'd do a lot bread boarding of circuits, of course, before we would finally package up models that could be taken out and tested in a system. But we didn't really have a research effort. It was a very low-level engineering design effort in circuits. Oh, we developed things. I developed a circuit for sampling. The question was, what was the voltage at a particular instant? Even though it might have been a relatively high-impedance impedance source, you'd like to sample it for a very short time at a particular time. I remember developing a transformer device with two back-to-back diodes, which were normally open either way so the signals can go through them freely. But then when you put a pulse on the transformer, the diodes would come together, and they would connect at that point to a memory capacitor. Then you'd have stored the sample. So I remember sampling a waveform at a particular instant and making a permanent record of it, in this case the capacitance on a voltage. I have a patent on that. That concept is used all the time in sampling today, of course, in the digital transformation, you do sampling and then you convert it into digital signals. I don't know whether the modern solid-state microcircuit devices have the same analogous concept or not, but the problem is to sample without interfering with the signal. I remember working on that one, and that was a development that I don't even remember being used. I don't even remember why I did it.

Bryant:

Were you aware of information and new knowledge coming from Britain, from their laboratory efforts?

Sherwin:

No, I don't think we were. I don't remember any new information about the display-end of the system.

Bryant:

What library facilities did you have?

Sherwin:

Well, the MIT Library was right there, and we used that.

Bryant:

So you had full use of their facilities?

Sherwin:

Yes.

Bryant:

I guess you had your own library of classified documents or specialties?

Sherwin:

Yes, they had file cabinets full of reports and so forth.

Bryant:

Eventually you were on the Steering Committee?

Sherwin:

No. I was a co-group leader with Ted Soller. The person on the Steering Committee that represented our part was Leland Haworth, who stayed there during the war. He was there the whole time.

Participation of Engineers

Bryant:

What can you say about the social organization of the place? Engineers had to be brought in at some point to do work that they were more accustomed to doing.

Sherwin:

In the indicator group I'd say about half of the people were engineers right from the beginning.

Bryant:

From the beginning?

Sherwin:

Yes. One of our best engineers, Bob Walker, I discovered yesterday, had never had any formal education at all. Yet he was one of the most skillful circuit designers I ever saw. He had a habit. He would never, never connect up what we called the twiddle box. He'd design a circuit. There were vacuum tubes and there were voltages and passive resistors, and Walker would figure everything out. He would expect certain waveforms and certain voltages, and he would build them. Or a technician would build them. Then he would set it up and test it out. A lot of the other people had boxes. They had boxes with a whole bunch of resistors in them, and they would keep trying one resistor after the other [Chuckling] until they got what they wanted. We called these twiddle boxes. Walker would never use twiddle boxes. He would go back and figure out what needed to be done. Go back and measure the resistors to be sure they were on spec and so forth. Then he would change the design of the circuit, after having thought about it, until it worked right. So we had people that were really very practical engineers, without formal training, or very little formal training. We also had some very good technicians. Most of the technicians came from the Boston area.

Plan Position Indicator

Bryant:

What do you regard as your most important work at the Radiation Lab now?

Sherwin:

There are two items that sort of stand out: the fully electronic Plan Position Indicator. The problem was a competing method of making a Plan Position Indicator, which is making the mechanical coil rotate with a servo-control, and it produced very precise circles. But it was clumsy, heavy and rather elaborate. So we wanted to have an all-electronic display. We worked on that at great length, developing methods of getting proper modulation of the sweeps. One of the biggest problems was that you wouldn't get exact circles because of the deflection system. You say you take an X and Y each out to 1.4; well, the circle was not exactly at 2. It would be in or out a little bit so the circles were all slightly square. [Chuckling] But that had to do with the properties of the yokes, deflection yokes. Afterwards, I noticed looking at TV tubes, they have a very elaborate system of the display coils on them. They're very specially shaped in order to produce pictures where straight lines appear linear. So a lot was learned after that. We worked with some modifications of deflection yokes to try to improve them, but I don't remember making a big improvement. The other one was the display for the ground control approach, which was a method of magnification that transformed linear tracks into linear tracks, and still gives you lots of sensitivity in the vertical direction. And the development of that oscillator capacitor device to compensate for the nonlinear scan signal or output shaft. I worked on both of those myself.

Bryant:

Who did the off-center PPI?

Sherwin:

Several people worked on that, and I don't remember who. That was great. We were struggling with the limitations of the power levels and voltages of tubes in coils. So when you start going off-center, you now have to produce sweep signals that are even longer. You have to go way up and double the signal strength. You also had to be struggling constantly with defocusing problems. You'd start getting out to the edge of those tubes, you'd get defocusing, partly, again, due to deflection yokes. I think that that deflection yoke business was one of the major developments that was improved after Radiation Lab. I don't think there was much improvement during that time. There was no real scientific attack on how to improve the deflection coils on cathode-ray tubes. Now, looking back, we probably should have been working on it.

Bryant:

What companies were involved in manufacturing what you designed?

Sherwin:

General Electric was building the display systems for the SCR-598. That was a PPI. But I don't remember who built the equipment for the ground control approach. Oh, yes, I do. It was some company in Los Angeles because I remember going out to Los Angeles once on the GCA circuit, and the system was being built out there. I think it might have been North American, what later was Autonetics, building that. Because I remember flying out there in '42 or '43 in DC-3s that took 23 hours to get to Los Angeles from Boston.

Bryant:

Without an overnight stop. [Chuckling]

Sherwin:

Yes, there were no stops.

Influence on Television Technology

Bryant:


Audio File
MP3 Audio
(082 - sherwin - clip 1.mp3)


I've heard lots of people ask how much influence did the radar developments, techniques and technology of World War II advance the state-of-the-art for television, a starting point for television?

Sherwin:

Yes, I have some thoughts on that. The group that left the indicator group most of them went with General Precision Laboratories in Pleasantville, New York. Garman went there as the head. He set up a research lab for doing two things: improving TV display systems and also setting up new trainers. He had been in charge of the simulator group for training people on radar operations. This group went there and I heard later about what they got involved in. They carried on what we had started, which was superimposing more than one picture on the same screen, and putting systems off-center, and having two different displays on the same screen. They went ahead at Pleasantville later and developed much of the sophistication in the multiple-image displays that are on TV. They also did a lot of work on improving the precision of circuits and of deflecting coils and so forth to produce high quality across the entire screen. They also got involved in something I didn't know about this until recently; I had a talk with Clayton Washburn, one of the people in our group, who said they worked to greatly improve the resolution of the TV picture without changing the transmission signal due to the way in which the spots and the deflections were used on the cathode-ray tubes. He said that you could get a 50 percent improvement — maybe even 100 percent improvement — in your resolution without actually changing the bandwidth of the transmitter.

Bryant:

And definition?

Sherwin:

He said the biggest bottleneck was in the definition on the screen. It wasn't actually in the signals.

Bryant:

Is that right?

Sherwin:

In fact he's been spending a lot of time recently to try to convince Zenith, our last TV manufacturer, and he's also talked to the people who own the TV stations. He says, "Look, you're getting this high-definition TV beam coming in, and it's being driven by the Japanese. If you follow their track, you're going to have to change all your transmitters. You're going to have to change the number of channels. Because you need much wider bandwidth." He says, "You can gain almost all of that without making any changes at all, by just changing the way the displays are made." I'd never heard this before. He has been carrying this on for a number of years. So there was a lot of improvement. The group from the Radiation Lab, I think, made a major impact on the cathode-ray display systems that were developed over the next five to ten years and came out commercially.

Bryant:

That's a very major item not only because of television, but because so many things we have now are cathode-ray displays. We don't have anything that exceeds it in efficiency or cost.

Sherwin:

Some of the flat-screen displays are beginning to compete. But they're not as fast as the cathode-ray tube. Washburn said yesterday that it's surprising how you always promise you're going to get all these totally new technologies for displays, and then we come along and improve the cathode-ray tube again and up goes its performance. It's very flexible and very inexpensive, really. So the people that went there had a big influence later in rapidly improving TV, though I'm sure that those improvements would have been invented anyway. But they were invented clearly more quickly because of the experience and training of these people that went there.

Bryant:

One-to-one correspondence.

Sherwin:

Yes, yes.

Control Systems Lab

Bryant:

Correlation. Are there any other civilian outcomes of the efforts at Radiation Lab?

Sherwin:

In the areas in which I had any experience?

Bryant:

Yes.

Sherwin:

I don't remember any. There were things at Radiation Lab that I was aware of there and which I heard discussed yesterday having to do with the Doppler properties of signals that are observed in airplanes, which led to the very high-resolution radar systems. In fact yesterday at Britton Chance's group, they talked about the technique they discovered to determine the ground track of an airplane, a very important piece of information to know in a bombing. They were developing bombsights, essentially. You need to know the actual motion of the airplane on the ground. It turns out that if you have a beam, which is typically 3 or 4-degree wide, and you run it along slowly, when it's off the ground track the intensity is constantly dancing around. You see the little scintillations on the scope. But when it hits the ground track, the scintillations disappear. Well, those scintillations are the Doppler beats between the various targets that have different radial velocities. That is the heart of the synthetic aperture radar system. Instead of beating against each other, if you beat the Doppler signals against a reference signal in the airplane, you now have a precision angle measurement device. This is in a high-resolution radar system. See, the Korean War came in 1950. We formed a University of Illinois laboratory called the Control Systems Lab (Wheeler Loomis set it up originally), and we had about 30, 40 people. We carried on everything in the same tradition of the Radiation Lab.

Bryant:

So that was another result from the Radiation Laboratory experience.

Sherwin:

Definitely was a result. The same management, many of the same people, and the same technique. The same idea was that the scientists (physicists) in this case (and some engineers again) were exposed to the military requirements. In other words, they would go to lectures and series. They would have seminars. I remember we attended one on tactical aircraft control and tactical air defense. We went out to Cal Tech, and they had a three or four days' series at which a lot of military people presented the problems of tactical air defense. One of the projects chosen by the Control Systems Lab was developing tactical air defense systems. So they started to work on it, and they were now with computers. We had the ILLIAC I there, too, one of the early digital computers and they were starting to use computers keeping track of multiple targets. The basic military requirement was stated by the users. Then the approach and the systems design concept was worked up by the engineers and physicists. And then it was demonstrated. Once it was demonstrated in a working model, a sort of prototype model, then the military people could appreciate whether it did or did not meet the requirements. So that cycle of exposing the scientific people to the application problem; then giving them the resources on their own to meet the need; building some sort of a prototype; and then bringing the applications people back and saying: Here's what your problem was; here's how we think it can be solved, and we've done all this work; it might be a year's worth of work; you come in, and then you decide whether it's worth going on. That cycle was used again and again.

Bryant:

So that was used again and again in the Radiation Laboratory?

Sherwin:

That's right. It was the same cycle as the Radiation Laboratory, and it was applied at the Control Systems Lab.

Bryant:

The Control Systems Lab at the University of Illinois was established in about 1955?

Sherwin:

1950 or '51. I don't know which. It was established almost immediately upon the beginning of the Korean War.

Bryant:

I see. And Wheeler Loomis was the initiator?

Sherwin:

He was there. He was the head of the Physics Department, so he went over and started the Control Systems Lab.

Bryant:

How interdisciplinary was the Lab?

Sherwin:

Physics, electrical engineering, and some mathematics. But mostly physics and electrical engineering.

Bryant:

And you were?

Sherwin:

I was professor of physics.

Bryant:

And you were associated with the Control Systems Lab in what capacity?

Sherwin:

I was a middle group leader there. I picked a project. I had done some physics work on neutrino detection in which I looked for missing neutrinos. But I probably had some good work in physics going there, and then this came along. So I dropped my research work in physics, and I went over to the other building where they'd set up this laboratory. There was an operation requirement for improving the moving targets, detecting trucks and people walking around on the ground. I got involved in that with a small group of about four or five people, trying to improve moving-target detection. We were trying to [detect] technical moving targets from an airplane, and in the process of doing it, we were flying the airplane. We were doing careful frequency analysis of signals received from radar at range intervals off-track from the airplane and we observed some very peculiar effects. A distinct frequency of signals would come and go as this patch of sensing at even a fixed position. Here's the beam. The airplane is going this way, and the beam is scanning along the ground. We discovered that different frequencies would come and go, and all of a sudden I realized (this was one of the things that I was involved in personally) that those were ground targets beating against each other. And then I said, "Oh! All we have to do is put a coherent transmitter in the airplane." We did, and it worked fine: good, high resolution pictures within the beam width.

Bryant:

Where did the idea for the Control Systems Lab originate? Do you think the Armed Services asked for the facility? Or did Wheeler Loomis and his associates have an idea that they would sell?

Sherwin:

I don't really know. But I think that Wheeler Loomis went to the Navy and the Army proposing that they set up what was called a Tri-Service Research Laboratory, which from the beginning was called a Tri-Service Laboratory.

Bryant:

Louis Ridenour was there as a dean of the graduate school.

Sherwin:

He was involved in that. Right.

Bryant:

So he would have been involved, too.

Sherwin:

Right. It was clear that the Radiation Lab system had been effective, and the idea was to duplicate it on a smaller scale at Illinois. Ridenour was there, Loomis was there, and they sold a contract to the military. I think each service put in a few hundred thousand dollars. It wasn't a big thing at the beginning, and the goal was to establish the same principle of exposing the scientists/engineers to the problem, having them go off and work out proposed solutions, demonstrate the feasibility, and then see if they were acceptable.

Bryant:

Very interesting. I had not heard that brought out, that the Control Systems Lab at Illinois would have been patterned on Rad Lab.

Sherwin:

It was patterned exactly on Radiation Lab.

Charles River Project

Bryant:

Did Wheeler Loomis have to do with the establishment of the Research Laboratory of Electronics?

Sherwin:

No, I don't think so. But he was involved in Lincoln Labs.

Bryant:

What was Lincoln Labs? What was that called?

Sherwin:

Well, Lincoln Labs was called the Charles River Project, and later became Lincoln Labs.

Bryant:

What was Wheeler Loomis's position in starting that Charles River Project?

Sherwin:

The project was a summer study. That was a very common thing, to give a summer study.

Bryant:

Oh, I see.

Sherwin:

Since he had been Associate Director at Radiation Lab he was asked to set up a summer study at Charles River Project at MIT. A number of people went there, including some of the people from Illinois; and they studied the air defense problem in general — the continental air defense problem. I think that Loomis and Fred Seitz were very interested in being sure that science was used for national defense, more so than many of the scientists are today. He and Seitz decided that they should have a lab at Illinois, and they should focus on different areas than the Charles River Project. The Charles River Project spawned the Lincoln Laboratory. The Lincoln Laboratory, as I remember, was focused upon continental air defense. That was their big project. Same pattern [as] at the Radiation Lab: overall function requirements/operation requirements given to the laboratory; the laboratory would work out partial solutions and simple working models; and then they would cycle back. Is it doing the job? Is it effective? And after there was an agreement, then it would go into some production. Lincoln Labs got IBM involved for the first time in a big way of using computers. In effect, that's the first time IBM really got into digital computers. Up until then, they were just the cardpunch company. Lincoln Laboratory launched IBM, in the modern computer sense, when they built the big computers that were used in the air civil defense jobs.

Bryant:

Because of the demands for the system?

Sherwin:

They had a special name for that computer [AN/FSQ-7], and I can't remember it. But it was the digital computer that kept track of all of the aircraft information in this network. That computer was built with transistors. It was the first really big transistor computer, and it was the thing that gave IBM the lead in building large computer systems. From there, they just kept on going up.

Synthetic Aperture Radar

Bryant:

There must have been other laboratories that could be traced back to the influence of the Radiation Laboratory, the people and their experience, as well as the type of organization.

Sherwin:

Well, Michigan. Project Wolverine occurred at the same time. They concentrated on battlefield surveillance projects.

Bryant:

They had Army support for that.

Sherwin:

They had Army support; it was not tri-service. But it was, again, built with the same principle the Radiation Lab had. Which was that the Army would describe their operation requirements, and the laboratory people would get involved in solving it.

Bryant:

What Radiation Laboratory people went there?

Sherwin:

I don't remember.

Bryant:

I can't recall any.

Sherwin:

But Wes Vivien was one of the early people there, and Lou Cutrona and others. I got off the topic a minute ago. We were concentrating on moving-target detection. Then we realized that with the synthetic aperture radar was possible, and you could do precision focusing with it. But this was always a side product from snorkel detection. We'd picked snorkel detection as our immediate goal. We went down to Key West to detect snorkels, and, sure enough, we had a sea spectrum. We could see the seawater spectrum, and over here was the snorkel displaced way out, separated out. And it had the proper speed from the sea clutter. Same radar! We put the beam out and ran along the coast of Key West, Florida, recording on our tape all the signals from that strip. We took that section back, put in the loop, went round and round and measured each frequency and adjusted it, and there was a map of Key West, Florida, all taken within the beam width of the radar. Wes Vivien pointed out later that was the first — what do you call it? Where you have a picture without an image on it and reconstruct the image?

Bryant:

You mean a holographic?

Sherwin:

Yes, a holograph. It was actually the first holographic image.

Bryant:

I guess it is, yes.

Sherwin:

Because what a holograph is, you store signals from a distant point against a reference in which you have the phase reference stored as well as the signal. Then you re-play it back to product the spatial picture. That's exactly what we did. We sent out coherent radar signals. We stored them against a reference signal, in phase, on a tape, and then we played them back in such a way that they produced a map. So it was really the first hologram. Then of course at that point, Project Michigan was formed. I think it was the first Wolverine study at Michigan. They got very interested in it, and they made very big improvements in it. They went ahead and built the system that actually worked with the synthetic aperture radar system and had it working by 1958. From there it's gone way, way up. Did you happen to see Ivan Getting's lecture yesterday?

Bryant:

Yes.

Sherwin:

He had some pictures there of the Detroit River taken with a modern synthetic aperture radar, and it's quite incredible.

Bryant:

I've attended two or three of their symposia on synthetic aperture radar. They used to have those symposia regularly. You've attended some of those, too?

Sherwin:

I was at one of them.

Bryant:

The first display I ever saw with synthetic aperture radar, I just almost jumped. It looks like it's coming out at you.

Sherwin:

I know, I know. That whole thing came out of the exact pattern of the Radiation Lab, in which the problem was ground surveillance, in this case originally in submarines (periscope surveillance) in which the engineering people had the authority to design equipment and test it. The Air Force supplied a C-46. We put the radar in it and put our echo box in to get coherence, and put our tape recorders in. And we went off to detect snorkels in Key West. We weren't working with any formal specifications from the military for that. We had general specifications.

Bryant:

You had a requirement in mind.

Sherwin:

We had a requirement very well in mind. But it was not specified in all of the special, elaborate form they now have.

Bryant:

I think that's very important.

Sherwin:

Yes. This is still happening today. In our local paper was a story about a little company in San Diego that has been for years building off-road buggies by modifying the VW dune buggies. They put on bigger wheels and sturdier suspensions and special filters to keep out dust. They can go anywhere. They can run right through sand. They've been selling them to sports enthusiasts. They are so good, they can go 80, 90 miles an hour across a desert. So what happened was the Army got so interested in them, and they ordered some. This guy only had a company with three people in it, and once every few days they'd make one more. They brought in all his friends, and they set this thing up, and they made about 30 of them, all of which are hauled over to the Gulf War on an airplane. They ran all around for reconnaissance, all around in Kuwait and southern Iraq, and the Iraqis couldn't touch them. There were no specifications. What they had taken was something that was designed, so they used it. Another case happened. I don't know if you noticed in Gettings talk, he said they had all these commercial global GPS systems. Receivers were in commercial production. But the Army had had such an elaborate specification for field-operated, hand-held, global-position satellite locators, that they didn't exist. When they went over to the gulf they didn't have them. They were still going through the system of formal R&D and production. It takes five, eight years for something to get produced. So they went out and bought these commercial ones, and they hauled them over to Kuwait.

Bryant:

Yes. There were two channels: one, the military channel; you get a lot more accuracy than you do out of the civilian.

Sherwin:

Yes. But all they had was a civilian channel, but they knew within 30 feet of where they were — a couple hundred feet. The point was that the civilian application device actually was very useful militarily.

Bryant:

Yes.

Sherwin:

They bought all they could get [Chuckling] without having written the specifications.

Bryant:

You mentioned the use of the echo box. Echo box, I know, was used in connection with beacons.

Sherwin:

Right. That's where we learned about it.

Bryant:

Technically what was the actual function of that?

Sherwin:

The echo box was a method of remembering the phase of a transmission. You remember magnetrons. Every time you start them off, they start off at a random phase. So from pulse to pulse there's no relationship. If you take a magnetron and take a tiny fraction of its power and put it into a very highly resonant cavity, you'll get an exponentially decaying echo as the signal goes back and forth in the lowest loss mode. I think it's a very simple coaxial cylinder. That echo is like having a very long pulse. If your signal coming back from the distant target comes back before the echo has died away, you have, in fact, reserved the phase of the transmitted echo.

Bryant:

So it's a high enough "Q" that even after 20 or 30 or 40 microseconds you've still got signals?

Sherwin:

Yes. You could look out 10, 20 miles.

Bryant:

Would that have not been useful in the Moving Target Indicator? All the effort to make mercury-delay lines?

Sherwin:

It would have been, but for some reason at Radiation Lab it wasn't. It would have been a very good way to get the relative motion of the aircraft with the distant target. Moving Target Indicators depend upon Doppler beats between the target and the ground around it, which was standing still. The ground provided a coherent reference, so to speak, for each pulse. But the echo box in the transmitter allows you to measure the speed of the aircraft with respect to the target. That was not used during Radiation Lab, as far as I know for MIT.

Bryant:

Interesting.

Sherwin:

It was first used only when we started to use it for snorkel detection.

Bryant:

Uh huh. Later.

Sherwin:

Later. That was five or six years later.

Simulators and the Training Group

Bryant:

Another thing. You mentioned the simulator group. Was that in existence when you came? And why was it set up?

Sherwin:

It was called Training.

Bryant:

Training?

Sherwin:

It was called Training. They were concerned with models, and I wasn't too familiar with them. But they were concerned with training people to run radar systems. So they would have complete radar systems that would simulate echoes and tracking and plotting.

Bryant:

Well, they had to have displays.

Sherwin:

They had to have the displays for those. They used sort of standard displays.

Bryant:

They used standard displays for that?

Sherwin:

Yes. I don't remember what they used for modeling. Whether they had ultrasound tanks or not. I don't remember that. That was used later I know. Where they had a sound pulse.

Bryant:

A transmission path of some kind.

Sherwin:

And they had a model with high-frequency sound in a liquid. I don't remember that, but I do know they had a training group. The training group became later one of the leading groups in synthesizing displays that would duplicate operational conditions for military applications. I think probably later for commercial aircraft flying, where they used radar systems for navigation.

Postwar Career

Bryant:

What effects did your Radiation Laboratory experience have on your subsequent career?

Sherwin:

It didn't have a very good effect. The trouble was that I was motivated emotionally and intellectually to go into pure physics, and I regarded the Radiation Lab engineering work as a digression and a secondary level of quality. The same thing happened when I got to the Control Systems Lab, that I was deflected from my true goals. It really did have an effect on me because by the time the Control Systems Lab ended (I got out of it in 1955) — so for about a 15-year period — I'd spent over half my time doing engineering and not physics. It really got me off the track, and I got behind in physics, which I never really caught up with. So it deflected me. I did some very good work in engineering, which I was not very proud of. For example, I worked out some of the theory of signal detection, which later I found was published by somebody else, and I could have published it. I made the first synthetic aperture radar picture, but I never made a patent application. I just didn't feel that engineering was important.

Bryant:

So you were making contributions to engineering?

Sherwin:

Yes. I was making more contributions to engineering than I was to physics.

Bryant:

You were solving problems.

Sherwin:

Yes, yes. I enjoyed engineering and later got into a lot of it.

Bryant:

Well, at Illinois you had a stint with the Air Force Command?

Sherwin:

Yes. I was Chief Scientist at the Air Force for a year. They had a rotating job there. I worked with Jimmy Doolittle, which was a great experience. He's a fine person.

Bryant:

Was that a Pentagon job?

Sherwin:

In the Pentagon, right. As a matter of fact, it was my initiation of the program that ended up as the big Early Warning Radar Systems that existed at Thule. They first built gigantic rotating mechanical beams to pick up missiles for early warning. I initiated that in 1955, and it got built within three or four years. Getting showed pictures of the modern Thule radars, those great big flat-faced ones.

Bryant:

Phased array?

Sherwin:

Phased array things that could track a thousand targets at once. That was the next generation. I contributed to the earlier design. It certainly was engineering, and I kept getting further and further away from physics. I tried to get into physics again, and I got interested in writing some books. I'm really fascinated with the problem of trying to simplify and organize knowledge so that it's easier to understand. I wrote a book on Basic Concepts of Physics to try to compress things in a very simple way. Also wrote a book on Introduction to Quantum Mechanics, which was very widely appreciated by the students but not by the professors. The students loved it because it was very clear. In 1960 I left to go up in administration and never really got back to physics seriously until recently, when I started up my work in relativity at home, after I retired.

Bryant:

You were at Aerospace Corporation?

Sherwin:

Yes. For approximately three years I was head of the laboratories.

Bryant:

Aerospace's work is air defense?

Sherwin:

No, they were systems engineers for offense systems and for space systems. As a matter of fact, the Global Positioning System, which Ivan Getting talked about, that's his baby. Ivan Getting's. He was the one that was convinced that that system had all the properties for navigation worldwide that was needed. Much of it was carried through at his insistence. Indeed it has turned out to be one of the really great space systems. Really a great thing.

Management Model of Rad Lab

Bryant:


Audio File
MP3 Audio
(082 - sherwin - clip 2.mp3)


We've just hit some high spots here. Are there some things you'd like to talk about? Summarize or talk about? Such a wide variety of activities took place. So many people were involved. We've seen it expand in so many ways as a result of the influence there.

Sherwin:

Well, I think one of the consequences of the Rad Lab and then later the Tri-Service Electronics Labs was to provide a model for what later became the Materials Research Laboratories that were funded through ARPA, I think, mainly. They had the same property. The universities would set up a research center in a field of interest to the Defense Department. Then they would be given stable funding and local control of initiation of projects without having to have a lot of prior activity reports on it. If they were doing a good job, their funding would be continued. I think it led to the model of the institutionally managed research projects. This is very much opposed to the desires of most of the scientists, who want to have it controlled by committees and peer review. They want to have project-by-project review by a peer [group] of outside scientists who then say whether NSF or DOD or some other government organization can finance some particular project. It's called peer review control.

The other picture is the Radiation Lab, the Control Systems Lab and the Materials Lab. The control is in the director. If you want to work on a project, all you have to do is convince the director that you need to have $10,000 to do this or a hundred thousand, sometimes, to do that. And if he's got the money, he's got the authority to use it. So there's a real clash of cultures between the management on the part of the university, in this case, for the control in the hands of the professors. They much prefer the professors, particularly the ones that are on all the committees and make all the decisions. The problem is that universities in general do not have any command organization; there's no authoritative control. All department heads ever do is hire and maybe sometimes fire people. They don't really control projects. The university doesn't trust local management. They trust only their peers to review whether what they're doing is good or bad. So it has produced a real clash and, I think, a very desirable one. It's shown that if you give a local university the opportunity to have a good manager of, say, a materials lab, that you get first-rate research out of it without having a bunch of peer-controlled projects. You get much better interdisciplinary cooperation; you get all sorts of advantages. I think that Radiation Lab has encouraged management-control R&D by skilled managers who were themselves competent researchers. One of the main things I did when I was at DDR&E (I was responsible for research and technology) I insisted that they establish for every one of the in-house laboratories a budget of 10 percent of the total budget available to the director of the laboratory for the initiation of projects which he felt would be of value to the supporting services. These did not have to receive prior control. Most of their work was done in support of existing systems and to requirements. But he had this initiation of a certain percentage. A number of very brilliant things were done with those funds by those laboratories, controlled by the directors, the local director of the laboratory.

Bryant:

I was just thinking of the DDR&E. That's your time in the Pentagon?

Sherwin:

Right. I worked with Harold Brown there. He recruited me. He was the Director of the Defense Research & Engineering, which is now called Under Secretary for Research & Engineering. But he was the Director of Research & Engineering, and he had four or five deputies, of which I was one. I took the Research and what was called Exploratory Development or Technology part of the budget.

Bryant:

You didn't do a stint in the Commerce Department as well, did you?

Sherwin:

Yes, I went to Commerce for about nine months after I left DDR&E.

Bryant:

In what part of it?

Sherwin:

I was the Deputy Assistant Secretary for Science & Technology. It was a poorly defined function. I didn't do much of any value. Herb Hollomon was the Assistant Secretary for Science & Technology, or something, at Commerce.

Bryant:

Who was the Commerce Secretary then?

Sherwin:

John O'Connor. This was John T. O'Connor, who was Secretary of Commerce under Pres. Lyndon Johnson, and who was formerly CEO of Merck and later CEO of Allied Chemical. That was a very disappointing experience in my career. One of the lower points, I must say, looking back at it. It was very unrewarding.

Impact on Education

Bryant:

There are such dead-end projects, I'm sure. Well, you were involved in teaching, both at Columbia and Illinois. Do you have any comment on the influence of organizations like the Radiation Laboratory and the World War II experience on education? Or teaching and research?

Sherwin:

No, I just don't think I do. They didn't impact education, to my observation, the methods of education or the substance of it even.

Bryant:

What about engineering? It would seem to me like post-World War II engineering got a lot more scientific and a lot more mathematical than it ever had before.

Sherwin:

I wasn't really ever involved in teaching engineers.

Bryant:

You didn't have that observation on engineering.

Sherwin:

Yes. But it had little effect on physics. It had an effect on physics in the instrumentation sense. People developed methods of instrumentation, which improved it. My first research project in physics at Illinois was used time-of-flight systems for detecting the speed of recoil particles, all of which was right out of the work I'd done in the indicator group at Radiation Lab. So it affected instrumentation, but it didn't affect the teaching, or didn't affect the ideas.

Bryant:

I wouldn't think so.

Sherwin:

But in engineering it would well have had an effect.

Bryant:

Well, I thank you very much for your time and doing this.

Sherwin:

Well, it's been very enjoyable. [pause]

Project Hindsight

Bryant:

Go ahead and introduce the subject.

Sherwin:

Okay. The subject is Project Hindsight, which I was involved in when I was at DDR&E. The objective was to look back and compare one military system with a successor system, typically ten to fifteen years later, which had different properties, and to try to see what innovations made possible a successor system. Ivan Getting was talking about one dramatic case, the original 584 radar. He showed the modern radars that can track a thousand targets with the precision one hundred times better, with a range accuracy a hundred times better, etc. The question is, what were those innovations in between? And more important, where did they come from? What motivated them? I had an interview with David Van Keuren from Naval Research Lab about that. One thing that I concluded was that when it came to applied science, it was very important that the local laboratory management have control of detailed projects, rather than working in response to case-by-case specific projects out of a central office.

Project Hindsight made a study, and they found that practically all the innovations that made a difference in weapon systems were motivated by people who had direct access to the military requirements. There were some that weren't, but they were usually applied requirements in the civilian sector. I remember being recruited by Shockley at the end of the war and talking to me, he said, "We're working on trying to control currents in solid state because we want to build a better switch for our switch." This work was applied, and of course they developed the transistor, which produced these great revolutions. But it was motivated by a very well defined operational need. So when you're interested in closing the gap, that's the way to do it. If you're not interested in closing the gap, and you just want to keep the scientists happy, you just let the peer review people decide what's the most exciting science. Now that's not bad; that's good. But it doesn't produce results that are of practical use in a reasonable time. The basic science ideas tend to stew for 20 to 50 years before they really have an impact. The ideas that come from an exposed need have an application in 10 to 20 years.

Gulf War

Bryant:

If you've got the time, there may be something from this recent Gulf conflict, which certainly was an application of science and technology to military needs, like skills we hadn't had before.

Sherwin:

It was dramatic! I couldn't believe it was happening, but I was told by the Chief Scientist of NATO about two years ago that if we ever got into a war with Russia, they wouldn't know what hit them. They wouldn't be able to run any of their radars after the first few hours. We would attack them. We always have this picture of these hordes of Soviets overcoming, rushing through Western Europe just like the Germans did. He said, "That wouldn't happen." The Russians would have been no better off than the Iraqis. Their equipment was incapable of stopping the destruction of their communications and their radar systems. They would be blinded in the first few hours of the raid, and they'd never get back.

Bryant:

You're talking about a technical superiority.

Sherwin:

We had superb technical superiority. Superb! Even against the Russians. The Iraqis are well equipped. They didn't have very good morale, obviously. But they were well equipped. But their military capability was just knocked flat in a matter of a few days. This was what this man from NATO said, "That's exactly what would happen in NATO if we had a war in Europe."

Bryant:

But doesn't that put the imperative on continuing to support both developments like this and offense?

Sherwin:

Right. He said that superiority of technology and good esprit de corps of the people and skill absolutely dominates. It's far better to have a small, well-equipped military force with the very latest and most efficient equipment, than it is to have a big one. So I think there's no doubt about it. Technological superiority is even more important than it used to be.

Bryant:

But unless you keep improving the equipment, the tools for these people to use, someday someone is going to move into new scientific developments/improvements that leapfrog it.

Warfare and Democracies

Sherwin:

I am an optimist in this regard. My study of war has shown that of all the wars that have happened since 1815 at the end of the Napoleonic Era, there have been 140 different wars with over 50,000 people involved on each side — out of those 140 wars, not one of those wars has ever been fought between two democratically governed countries. Not one. Britain fought Argentina over the Falklands. Argentina was a dictatorship. Britain didn't fight Iceland. They almost did. They had a big fishing argument with Iceland, but they wouldn't go to war with Iceland. They never would. Italy started off in World War I on the side of the Austrians and the Germans, but they were a democratic country at the time, and they switched sides and they joined the Western Allies.

Bryant:

They finally did, yes.

Sherwin:

Yes. So the story is that the growth of democracy — real democracy — has yet to produce a war between two countries so governed. There have been plenty of wars, but the wars are usually between countries neither of which are democracies. Democracies do get into fights, but with dictatorship countries. So I'm optimistic. I would like to see the Defense Department budget greatly reduced. The total forces go down, but the effort on superb instruments and equipment should continue full blast. We could have a military force half as big as it is now with even much better equipment than we have now.

Bryant:

Without having any inside knowledge, my observation is that Dick Cheney and the present military people would probably agree with that. They seem to be wanting to reduce the force but keep the support for development. They seem to be having trouble even getting money for Stealth, which I think was one of the most telling things to come out of that. Really, we've had Stealth technology for ten years and more, but it first convincingly demonstrated skill in Iraq.

Sherwin:

Our superb navigation in that trackless desert with these little hand-held satellite receivers was a tremendous military advantage.

Bryant:

But look how far you can take it? You can put those inside those cruise missiles.

Sherwin:

Absolutely right!

Bryant:

You can launch them from any place; start them out in any direction.

Sherwin:

You can get them within 10 feet of where you want them.

Bryant:

You don't have to track over a land mass first.

Sherwin:

Right. You don't have to use on-board radar. You don't have to depend upon a navigation-integrating device.

Bryant:

I could name you dozens of things that you probably could use that.

Sherwin:

I know, I know.

Bryant:

My field, if any, is guidance.

Sherwin:

Oh, yes.

Bryant:

[Chuckling] In radar that I've been associated with, the accuracy of guidance has improved so much. It's fantastic. You have these smart weapons. You do without the influence fuses, the tracking is so great.

Sherwin:

Yes. You can go right down the air vent if you want to, in the bunker. That guidance has really revolutionized warfare and made it possible to use relatively small explosives and have enormous effects, even on hardened targets.

Bryant:

In some cases you can even ignore the explosives, just rely on the impact.

Sherwin:

Yes, yes.

Bryant:

Which makes the weapons pretty small.